The Development and Applications of an e-PBL Platform

• Main Authors: Hu, Li-Ling, 胡莉玲
• Other Authors: Tseng, Shian-Shyong
• Format: Others
• Language: en_US
• Published: 2016
• Online Access: http://ndltd.ncl.edu.tw/handle/03980314306650134437

博士 === 亞洲大學 === 資訊工程學系 === 104 === In the capstone project, the project-based learning (PBL) is a popular strategy used by teachers to guide students to solve problem collaboratively and enhance students’ creative and problem solving abilities. However, the traditional PBL strategy cannot effectively enhance the students’ collaborative design and problem solving abilities without utilizing domain knowledge and previous user experience. In this dissertation, we propose an e-PBL platform based upon the enhanced Case-Based Reasoning (e-CBR) to help the students to complete the assigned project. The e-CBR consisting of cases retrieve, reuse, revise and retain can be used to reuse past cases and analyze the portfolios of similar cases, and the results can be used as scaffoldings to enhance students’ creativity and problem solving abilities in the capstone project. In this platform, the similarity-based case retrieval algorithm is proposed for retrieving the similar cases according to the project characteristic and the team profile, and the learning scaffolding generation algorithm is also proposed to find the longest common pattern between the given portfolio and those of the retrieved similar cases based upon our defined similarity functions. To evaluate the performance of the platform, the applications on collaborative design and storytelling-based game design have been done in this dissertation. According to the results of application on collaborative design, we have that the learning scaffolding can inspire students' creativity and brainstorming, sharing and reusing similar case can build team consensus. According to the results of application on storytelling-based game design, we have that the interaction dialogues can guide students to learn how to reason out the solution based on the backward reasoning strategy.